discovering-abstract-portal
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security risks were detected in the skill instructions or configuration files.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing external content retrieved from the AGW Portal. Ingestion points: External payloads from agw portal commands in SKILL.md. Boundary markers: Present; operating rules explicitly state to treat Portal payloads as untrusted. Capability inventory: Tool access is scoped to data discovery and inspection commands defined in SKILL.md. Sanitization: Present; instructions guide the agent to use field filtering and pagination to limit ingested data.
Audit Metadata